Activate for: ad copy, copy, landing page, subject line, headline, CTA, call to action, button text, email subject, A/B test, variant, tagline, value proposition, homepage copy, hero copy, meta description, Google ad, LinkedIn ad, Facebook ad, banner, paid social, copy variants. NOT for: long-form content (use content-creation), outreach messages (use outreach), campaign strategy (use campaign-planning), content scheduling (use content-calendar).
Scanned 5/27/2026
Install to Claude Code
npx -y skills add panaversity/agentfactory-business-plugins --skill copywriting --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Copywriting?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/panaversity-copywriting)More formats (shields.io, HTML) on the badges page.
---
name: copywriting
description: >
Activate for: ad copy, copy, landing page, subject line, headline,
CTA, call to action, button text, email subject, A/B test, variant,
tagline, value proposition, homepage copy, hero copy, meta description,
Google ad, LinkedIn ad, Facebook ad, banner, paid social, copy variants.
NOT for: long-form content (use content-creation), outreach messages (use outreach), campaign strategy (use campaign-planning), content scheduling (use content-calendar).
---
## COPYWRITING WORKFLOW
### Phase 1: Brief Collection
1. Copy type (see below)
2. Product / offer being promoted
3. Target persona (load from sales-marketing.local.md)
4. Core message: the ONE thing the copy must communicate
5. CTA: what action should the reader take?
6. Constraints: character limits, format, platform specifications
7. Brand voice (load from sales-marketing.local.md)
8. Number of variants required (for A/B testing)
### Phase 2: Copy Frameworks (select by format)
EMAIL SUBJECT LINES:
Framework 1 -- Specificity: "[Specific outcome] for [Specific persona]"
Framework 2 -- Curiosity: "[Unexpected claim or counterintuitive truth]"
Framework 3 -- Relevance: "[Company/role] + [their current situation]"
Framework 4 -- Urgency (use sparingly): "[Time-bound reason to act]"
Framework 5 -- Question: "[Question your prospect is already asking]"
Character limit: 50 characters for mobile display
Always produce: 8 variants minimum for A/B testing
Always include: 2 variants from each framework (mix approaches)
AD COPY (LinkedIn):
Headline: Max 150 characters | Pain or outcome focused | No jargon
Body: Max 600 characters | Problem -> Solution -> Proof -> CTA
CTA button: [Learn More / Download / Register / Get Started / Contact Us]
Thought Leader Ad (personal profile):
Opens like a LinkedIn post -- not an ad. Conversational. First-person.
300-600 characters. Ends with subtle link or CTA in last line.
LANDING PAGE HERO COPY:
H1: Outcome statement -- what the visitor gets/becomes
Not "Welcome to [Company]" -- always outcome-first
H2/subhead: Specific, supporting detail (who it's for, what it does)
CTA button: [Specific action, not generic]
Good: "Get my free assessment" / "Download the guide"
Bad: "Submit" / "Click here" / "Learn more"
VALUE PROPOSITION (positioning statement):
Format: "For [target persona] who [have this problem / want this outcome],
[Your Company] is the [category] that [core differentiator]
unlike [alternative], [your product] [key proof point]."
### Phase 3: Copy Variant Generation
For A/B testing, always generate variants that test different ANGLES,
not just different words for the same angle.
Test dimensions:
- Pain vs. gain (lead with problem vs. lead with outcome)
- Specific vs. broad (a named example vs. a general claim)
- Rational vs. emotional (ROI vs. risk/fear vs. aspiration)
- Short vs. long (within format constraints)
- Question vs. statement
### Copy Output Format
COPY VARIANTS -- [Type] | [Campaign / Context]
---
VARIANT A -- [Framework / Angle]
[Copy text]
Character count: [N]
VARIANT B -- [Framework / Angle]
[Copy text]
Character count: [N]
[... all variants ...]
---
TESTING RECOMMENDATION:
Lead test: A vs. B (pain vs. gain angle)
Expected winner: [A / B] based on [reasoning]
Metric to measure: [Open rate / CTR / Conversion rate]
Minimum sample: [N impressions / sends] before declaring winner
## NEVER DO THESE
- NEVER use banned jargon words (see global router: Law 5 list)
- NEVER use "click here" as CTA button text -- always describe the action
- NEVER produce only one variant -- A/B testing requires at least 2
- NEVER exceed platform character limits -- always check before finalising
- NEVER make a claim in copy that cannot be substantiated if challenged
- NEVER generate copy without loading the brand voice first
No comments yet. Be the first to comment!